Phenotypic heterogeneity and instability of human ovarian tumor-initiating cells

Abstract

The cancer stem cell (CSC) model proposes that tumors have a hierarchical organization in which only some cells indefinitely self-renew and thereby sustain tumor growth. In addition, the CSC model requires that tumor-initiating cells (TICs) be prospectively isolatable on the basis of their phenotype. Previous studies have suggested that serous ovarian cancer (SOC) conforms to the CSC model, but these used arguably nonfidelitous immortalized cell lines, cultured primary cells, or passaged xenografts as the source of tumor cells. We developed a robust assay for quantifying TICs from primary SOC. Using this assay, we find that TICs are rare when assayed in either NOD/SCID or NOD/SCID/IL2Rγ −/− (NSG) mice. TIC frequency (TICf) varies substantially between patients, although it is similar in primary ovarian masses and omental metastases, suggesting that TICf is an intrinsic property of ovarian tumors. CD133 marks all TICs from several primary SOC cases. However, in other cases, substantial TIC activity is found in both the CD133 + and CD133 − fractions, whereas still other cases have exclusively CD133 − TICs. Furthermore, the TIC phenotype can change in xenografts: primary tumors in which all TICs are CD133 + can give rise to xenografts that contain substantial numbers of CD133 − TICs. Our results highlight the need for quantitative rigor in the evaluation of TICs and for caution when using passaged xenografts for such studies. Furthermore, although our data suggest that SOC conforms to the CSC hypothesis, the heterogeneity of the TIC phenotype may complicate its clinical application.

Cancer stem cell definitions and terminology: the devil is in the details

“…In addition, perturbations of differentiation that affect normal lineage restriction processes (for example, as found in biphenotypic leukaemias), or the production of abnormal phenotypes, are well documented consequences of leukaemogenesis 41 . Similar examples of coexisting or unstable CSC phenotypes have also been reported in human ALL 42 and several types of human solid tumours 36,[43][44][45] . These findings highlight the fact that malignant cells have perturbed gene expression programmes that might affect the normal stability or the developmental control of the expression of certain cell surface markers, particularly when the cells are put in vitro.…”

“…This approach was pioneered with samples of cells from patients with various types of leukaemia (acute lymphoid leukaemia (ALL) 31 , acute myeloid leukaemia (AML) 32 and chronic myeloid leukaemia (CML) 33 ). Subsequently, the same approach has been successfully applied to solid tumours, including human cancers arising in the breast 34 , brain 14 , colon 16,35 , ovary 36 , lung 17 , and head and neck 37 . One criterion used to establish that these CSCs are biologically different from the bulk of the cells in the tumour has been the demonstration that they have a distinct phenotype that allows their prospective and selective enrichment.…”

Primary esophageal and gastro-esophageal junction cancer xenograft models: clinicopathological features and engraftment

“…Similar heterogeneity was seen in the corresponding xenograft, a phenomenon not typically described in cell lines. Unlike xenograft models of ovarian cancer 22 and adenoid cyst carcinoma, 23 the histology and morphology of most E/GEJ PTXG models remained relatively stable even up to late passages (P5 and beyond). Beyond these late passages areas of necrosis and cellular debris notably increased.…”

Epithelial ovarian cancer stem cells: underlying complexity of a simple paradigm

“…In a series of human EOC patient samples, the percentage of CD44 C cells within samples varied from 0.5 to 85% (Kryczek et al 2012). In addition, not all studies have found that CD44 expression defines a subpopulation of cells in primary patient samples that is enriched for TICs (Stewart et al 2011). It is also important to appreciate that most enrichment strategies based on the CD44 marker do not discriminate between the various isoforms.…”

“…In some instances, sufficient numbers of cells required for studying rare populations can be derived from primary patient material (Curley et al 2009, Stewart et al 2011, Kryczek et al 2012; however frequently this is not the case. Augmentation of patient-derived tumour cells can be achieved by the in vitro expansion of primary EOC cells as spheroids grown in low attachment tissue culture dishes, a method that has also been shown for enrichment of cells with stem-like properties (Bapat et Rather than reviewing the literature in a strictly chronological order, the contribution of the putative CSC markers to defining an ovarian CSC will be presented.…”